During the 2nd examination, a significant post-exercise increase in Treg count was observed solely in the placebo group, with subsequent normalization of this parameter after a 24-h recovery (Fig. 1a). Since lymphocytes are known to constantly migrate between the blood and other tissues, the post-exercise increase in circulating Tregs observed in the placebo group likely reflected a concomitant increase in the tissue count of these cells (Table 2). Previous research showed that Tregs may leave the circulation and migrate to lymph nodes and inflamed tissues [46, 47], whereby they mitigate the activity of cytotoxic cells and antigen presentation by dendritic cells (DCs) . This local suppressive effect of Tregs may exert an unfavorable effect on systemic immunity. Morgado et al.  demonstrated that an increase in the training loads of swimmers contributed to a significant reduction of cytokine synthesis by monocytes and dendritic cells. Immune impairment associated with heavy training loads may occur primarily in immunologically active tissues, at least partially explaining the difficulties in identification of accurate markers of immunosuppression that could be determined in peripheral blood. Published evidence suggests that the immune impairment caused by escalation of training loads has no specific laboratory profile [7, 50, 51]. It should be emphasized that circulating lymphocytes represent only 2% of lymphocyte population; the vast majority of lymphocytes can be found in immunologically active tissues, such as lymph nodes, spleen, intestines, blood marrow, thymus and skin . During heavy training, some tissue lymphocytes may migrate to the circulation in response to hormonal stimulation typical for strenuous exercise (e.g. catecholamines, cortisol). During post-exercise recovery, these lymphocytes migrate back to tissues, especially to those generating strong inflammatory/chemotactic signals . This hypothesis seems to be supported by the post-exercise increase in Treg count and subsequent post-recovery normalization of this parameter, observed during the 2nd examination in the placebo group (Fig. 1a).
ANOVA did not demonstrate a significant main effect of supplementation on Treg count; nevertheless, athletes from the supplemented group did not show a significant post-exercise increase in this parameter during the 2nd examination (Fig. 1a). Therefore, it can be assumed that supplementation with SPR played a role in the maintenance of lower Treg counts in tissues, preventing immunosuppressive effect of these cells and restoring an immune balance.
During the 2nd examination, athletes from the supplemented group presented with significantly lower pre-exercise and post-exercise values of Treg/CTL ratio compared to subjects from the placebo group (Fig. 2d). Lower values of Treg/CTL ratio in the supplemented group might also reflect a beneficial effect of SPR supplementation. The decrease in this parameter implies that SPR might mitigate the inhibitory effect of Tregs on CTLs. Due to lesser immune deficit, athletes from the supplemented group might have been better protected against opportunistic infections and reactivation of latent viral infections (e.g., with CMV, EBV and HSV-1). In turn, higher values of Treg/CTL ratio in the placebo group might reflect an unfavorable shift in the “overtraining threshold” associated with a radical deterioration of immunity.
Irrespective of the examination term and the study group, we observed a significant post-recovery increase in Treg/Tδγ ratio (Fig. 2c). This effect was modulated solely by exercise. Nevertheless, during the 2nd examination we observed a significant post-recovery decrease in Tδγ cell count in the placebo group, but not in the supplemented group (Fig. 1c). Due to presumable protective effect of SPR, preventing the post-recovery decrease in Tδγ cell count, strenuous exercise had probably less detrimental effect on the immune function in supplemented athletes. Tδγ cells play a key role in antibacterial, antiviral and antitumor immunity. Published evidence suggests that the number of Tδγ cells is inter alia modulated by exercise intensity. Anane et al.  demonstrated that either high- or low-intensity exercise stimulated an increase in the number of circulating Tδγ cells in previously untrained persons. However, another study showed that escalation of training loads during a winter training season resulted in a decrease in Tδγ cell counts in peripheral blood of elite swimmers .
Published data imply that SPR may exert a substantial effect on the activation of NK cells and their cytotoxic potential [35, 37, 43]. We did not observe a significant main effect of supplementation on either NK cell count (Fig. 1b) or Treg/NK ratio (Fig. 2b); however, irrespective of the examination term, strenuous exercise stimulated a significant increase in the number of NK cells (Fig. 1b).
According to a widely accepted concept, strenuous exercise stimulates migration of tissue lymphocytes to peripheral blood in a catecholamine-dependent mechanism. This hypothesis is inter alia supported by an increase in circulating lymphocyte count observed after infusion of epinephrine [55, 56]. However, the fact that the post-exercise increase in lymphocyte count in elite athletes was less evident than in untrained persons, implies that immune response of the former group to strenuous exercise may be weaker [57, 58].
Post-exercise changes in lymphocyte count in intensively trained athletes were markedly less evident than in untrained persons, and less pronounced than it could be expected based on epinephrine and cortisol concentrations. However, published evidence suggests that changes in hormonal parameters do not necessarily correlate with lesser mobilization of tissue lymphocytes and their migration to peripheral blood [8, 10]. Therefore, lesser responsiveness of tissue lymphocytes from elite athletes to strenuous exercise was interpreted as a consequence of training-induced changes, such as a decrease in the reactivity of ß2-adrenergic receptors on these cells, altered expression of adhesive molecules or their ligands [57, 58].
However, in our opinion, also other potential mechanisms contributing to reduced mobilization of tissue lymphocytes after strenuous exercise need to be considered. According to a widely accepted and empirically verified hypothesis, various stimuli (among them physical exercise and infections) may induce bidirectional flow of lymphocytes from/to tissues (spleen, muscles, lungs, bone marrow) . Adams et al.  showed that exercise promotes rapid migration of tissue lymphocytes (primarily from the spleen and lungs) to the circulation, along with markedly less evident shift of these cells from muscles to blood. Although this observation originates from a study in rats, the post-exercise increase in circulating lymphocyte counts in humans may also, to a certain degree, reflect higher absolute number of these cells in the spleen and lungs (and to a lesser extent, in muscles). Less evident post-exercise migration of lymphocytes to the peripheral blood of well-trained athletes may be associated with lower number of cytotoxic cells in their spleen and lungs, and/or with higher number of tissue Tregs. The latter hypothesis is also supported by the results of our study, which demonstrated a significant post-exercise increase in Treg count in the placebo group during the 2nd examination. However, also lymphopenia, observed few hours post-exercise as a consequence of the recirculation of blood lymphocytes to tissues (only < 10% of circulating lymphocytes undergo apoptosis), is a well-established phenomenon . During post-exercise recovery of well-trained athletes, their lymphocytes may primarily recirculate to the muscles; due to multiple training-induced microinjuries, muscle tissue synthesizes large volumes of cytokines which act as a chemoattractant for peripheral lymphocytes [59, 60].
According to Adams et al. , recirculation of lymphocytes from muscles to the blood is less pronounced and slower than between spleen or lungs and peripheral circulation. Probably, lymphocytes of well-trained athletes may remain anchored in muscles for a longer time, which makes them less prone to recirculation and more susceptible to apoptosis. This may result in permanent loss of some cytotoxic lymphocytes from the spleen and lungs. However, the exact reason behind a post-exercise increase in Treg count in well-trained athletes is still unclear. Perhaps, this phenomenon is associated with an increase in IL-2 (synthesized by T cells in response to repeated strenuous exercise), stabilization of Foxp3, longer survival of mature Tregs, and resultant increase in their number in some tissues [61, 62].
Our findings suggest that another, yet unidentified mechanism may exist behind the post-exercise immune impairment observed in elite athletes subjected to heavy training loads. Higher number of circulating Tregs may reflect their increased counts in immunologically active organs, presumably in the spleen and lungs. The number of these tissue Tregs that migrate to the circulation during strenuous exercise may be higher than the number of migratory cytotoxic T lymphocytes (other than NK cells). Furthermore, it cannot be excluded that immunologically active organs are not only abundant in Tregs, but also contain less cytotoxic lymphocytes (Tδγ, CTL). During strenuous exercise, these sparse cytotoxic cells may migrate to other tissues (primarily muscles) whereby they undergo apoptosis, which probably also contributes to a post-exercise immunity impairment.
Our present study showed, for the first time, that supplementation with SPR may modulate some components of the immune system in athletes exposed to repeated strenuous exercise. The fact that rowers from the supplemented group did not show a post-exercise increase in Treg count (Fig. 1a) implies that SPR may play a role in maintaining normal tissue level of these cells during strenuous exercise, thus preventing immunosuppression. Moreover, SPR seemed to attenuate a suppressive effect of Tregs on CTLs, since during the 2nd examination, athletes from the supplemented group presented with significantly lower pre-exercise and post-exercise values of Treg/CTL ratio than subjects from the placebo group. Finally, strenuous physical exercise did not exert a significant effect on Tδγ cell count in the supplemented group, whereas athletes from the placebo group showed a post-recovery decrease in this parameter. Altogether, these findings suggest that supplementation with SPR may exert a beneficial effect on selected components of the immune system in athletes exposed to heavy training loads.
Future studies should center around better understanding of the mechanisms of immune impairment activated during and after strenuous exercise. Another direction of future research should be the identification of factors that may counterbalance the unfavorable consequences of exposure to maximal training loads.